Improve ARM lowering for "icmp <2 x i64> eq".
authorEli Friedman <efriedma@codeaurora.org>
Tue, 18 Oct 2016 21:03:40 +0000 (21:03 +0000)
committerEli Friedman <efriedma@codeaurora.org>
Tue, 18 Oct 2016 21:03:40 +0000 (21:03 +0000)
commitc0a717ba5bf41ad90ba1ad8151f2b15b79c65998
treec8701a348c439c2afc62edc309a4a3fdcf791df6
parent36efa68463bc380c492cd08ae614c04ba3cd9fd7
Improve ARM lowering for "icmp <2 x i64> eq".

The custom lowering is pretty straightforward: basically, just AND
together the two halves of a <4 x i32> compare.

Differential Revision: https://reviews.llvm.org/D25713

llvm-svn: 284536
llvm/lib/Target/ARM/ARMISelLowering.cpp
llvm/test/CodeGen/ARM/vicmp-64.ll [new file with mode: 0644]